Re:Ask HN: 30y After 'On Lisp', PAIP etc., Is Lisp Still "Be...
by hello_computer on Wednesday June 05, 2024 @01:40AM
(#40581711)
types are just an example. structure (modules/classes/keywords/constructs) is also a big win for organization and assisted code navigation. even with dynamic types, i find it far easier to grep a python codebase than a lisp one. is it code? is it data? python syntax makes it fairly clear.
Re:Ask HN: 30y After 'On Lisp', PAIP etc., Is Lisp Still "Be...
by anonzzzies on Wednesday June 05, 2024 @01:53AM
(#40581770)
Fair enough; can’t say I have this experience but that’s fine; I am not religious. Although I would say, unless you are doing a lot of things with macros (hint; you shouldn’t and where you do, it should be clearly indicated; Ruby suffers from this as well when reading code imho) otherwise it seems very strange the code/data barrier isn’t immediately clear. But maybe I don’t hang as much on syntax as others do (we our own k3 to lisp to CUDA for some things and I don’t find it different/more difficult to read).
And we are pro static types, it’s just easier to add them after I am done experimenting. Not upfront. And we do via various (backward compat) custom CL constructs and provers.